A fully functional web application that simulates AI-powered debates on any topic using Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG). The platform generates structured, multi-perspective arguments (Pro and Con) with user interaction features like counterpoints, feedback, and scoring.
Enter a debate topic and watch AI generate compelling opening statements, rebuttals, and conclusions from both sides — with like/dislike feedback and vote counts.
zero-shot-ai-debate-bot/
├── backend/ # FastAPI Backend
│ ├── app/
│ │ ├── main.py
│ │ ├── routes/
│ │ │ └── debate.py
│ │ ├── schema/
│ │ │ └── debate.py
│ │ ├── services/
│ │ │ └── debate_engine.py
│ │ └── utils/
│ │ └── rag_wikipedia.py
│ └── requirements.txt
├── frontend/ # React + Vite + Tailwind Frontend
│ ├── src/
│ │ ├── components/
│ │ │ ├── Header.jsx
│ │ │ ├── TopicInput.jsx
│ │ │ ├── DebateDisplayArea.jsx
│ │ │ ├── ArgumentColumn.jsx
│ │ │ ├── ArgumentCard.jsx
│ │ │ ├── OverallVoteCounter.jsx
│ │ │ └── Footer.jsx
│ │ ├── App.jsx
│ │ └── main.jsx
│ └── vite.config.js
└── README.md
- ✅ Zero-Shot Topic Understanding
- ✅ Structured Debate Format
- ✅ Dual AI Personas (Pro vs. Con)
- ✅ Like/Dislike Feedback System
- ✅ Vote Counter Display
- ✅ FastAPI Backend with OpenRouter API
- ✅ Wikipedia Retrieval (RAG)
- Python 3.10+
- OpenRouter API Key
cd frontend
npm install
npm run devcd backend
python -m venv venv
source venv/bin/activate # or venv\Scripts\activate on Windows
pip install -r requirements.txt
uvicorn app.main:app --reloadRequest:
{
"topic": "Is AI dangerous?"
}Response:
{
"topic": "Is AI dangerous?",
"pro": {
"opening": "...",
"rebuttal": "...",
"closing": "..."
},
"con": {
"opening": "...",
"rebuttal": "...",
"closing": "..."
}
}- React + Vite
- Tailwind CSS
- Axios
- Lucide Icons
- FastAPI
- Pydantic
- OpenRouter / OpenAI API
- Wikipedia API (for RAG)
Team BotMinds
- Ishatva Singh Panwar
- Pratham Bajaj
- Hritvik Mohan
- Keith Kaur Malli
B.E., NITTE Meenakshi Institute of Technology



